Eugene,
More (silly) thoughts...

At first it seemed as if the symptoms could be related to a partially
plugged vent line (bugs, dirt, etc), which was perhaps aggravated by some
moisture.  However, it seems unlikely that both lines would be plugged and
that the results would be so binary at the same time.  Thus, I'd first focus
on the common components.

To test the vent line, open the filler cap and suck on the vent line (naca
end).  DON'T BLOW ON THE VENT LINE, AS ANY FOD MAY BE FORCED INTO THE TANK.
Also naturally, caution against inhaling any fuel and/or vapor.


What about an intermittent suction leak between the boost pump and the mech
fuel pump?  The mech pump would then just draw air (versus fuel).  You would
think that a loose fitting (etc) would also leak fuel - but I've seen poor
fittings that would suck air and yet not drip/spray liquid.

The mech fuel pump has a case seal (and associated drain).  If the case seal
fails, does/can the pump draw air???  If you run the boost pump with the
engine off and the mech pump case seal has failed, will fuel typically
run/drip out the drain? (that's my assumption as a test).  Do your injector
drains and fuel pump case drain all run together (making individual ID
difficult)?

Still unknown as to why it (seemingly) fixed itself and/or whether the
humidity had any relevance.  

At 12.5 GHP it sounds like you were LOP (guess).  Were you at similar power
settings on the later return trip when things were fine?

Rick